Duplex Sonography versus Angiography for Assessment of Femoropopliteal Arterial Disease in a “Real-World” Setting

Abstract

To evaluate the agreement of duplex ultrasound (DUS) versus digital subtraction angiography (DSA) for assessment of femoropopliteal arterial disease in a real-world clinical setting. Consecutive patients with peripheral artery disease who were scheduled for a percutaneous intervention were included in this retrospective study. During an 18-month period, 491 patients (276 men; median age 73 years, interquartile range 64-81) were enrolled. A peak systolic velocity ratio (PSVR)>2.4 was the optimal cutoff for detecting a >50% stenosis by DSA. Findings of preprocedural DUS in the proximal, middle, and distal ipsilateral superficial femoral artery and in the popliteal segment were analyzed for agreement with preprocedural femoropopliteal DSA using kappa statistics. Only the target limb in each patient was analyzed, for a total of 1964 vascular segments. Agreement for the degree of stenosis in 10% increments was only moderate (weighted kappa 0.67, 95% CI 0.65 to 0.69). Using the PSVR>2.4 cutoff, agreement between DUS and DSA for a >50% stenosis was good (kappa 0.79, 95% CI 0.77 to 0.81). Sensitivity, specificity, positive predictive value, and negative predictive value for correctly detecting a >50% stenosis by DUS were 0.81 (0.78 to 0.84), 0.93 (0.91 to 0.94), 0.84 (0.81 to 0.87), and 0.91 (0.87 to 0.95), respectively. Comparable findings were observed within different patient subgroups. Agreement between DUS and DSA in the femoropopliteal segment is only moderate with respect to the absolute degree of stenosis. However, detection of a >50% stenosis can be done with acceptable precision in routine clinical practice using PSVR>2.4 as a threshold.